Jump to content

Keep Hoping to Visit the Forum and See that UD bought the Insteon patents


Recommended Posts

Posted
11 minutes ago, RPerrault said:

as i understand it

one of the functions a plm performs is to insert data on the power line - the data being binary, so a 0 or a 1 - it does this by inserting the data bit on the electrical sine wave when it crosses the x axis - nothing new about that, others have done it - that topology is not a mesh - its more of a star topology since the data travels to your electrical breaker box to access devices on another circuit of the same leg - to traverse to devices on a different leg, i think the data travels to the transformer servicing your house and reenters on the other leg - and through all the homes services by that transformer - hence, the bridging devices needed for all the early insteon devices that did not support rf communication - some of us bitter old timers recall the lies smarthome led us to believe

that plm function, inserting data and monitoring the electrical sine wave, is something all insteon devices do - except battery operated devices i suppose - the specs for insteon allow for other devices to repeat the signal, but a provision is made to keep the repeating from continuing for 12 years (hop counts) - i am unsure if there is a provision in the insteon specs for collision detection and recovery (like the first ethernet specs) - my guess is that insteon has no routing function so every command sent on the powerline is a broadcast - all devices see it and only act on the data if it is addressed to them (0a:34:23 or whatever)

insteon also sends data via rf - that topology can be called a mesh - my question was 'if a device receives a command via rf, does it repeat that command via both rf and powerline?' - i suspect the answer is yes - because no one talks phase couplers since the disreputable smarthome began to implement both rf and powerline devices (like they lied to us by omission years ago) - if so, the plm function of inserting data on the powerline becomes moot - only a rf communication is needed to get the controlled broadcast storm started on the 'network'

another function of the plm seems to be - in order to have a controller - a link is established between the plm and each insteon device - since the insteon hub moved the plm functions into either the hub or the cloud, a plm can be eliminated and its functions handled by a controller - my guess is that the disreputable smarthome group placed setup and control functions issued bythe hub into the cloud - insteon will function without a controller or hub but automated stuff like timers and programmatic functions need the controller - that needs to issue commands to insteon devices - the with the built in rf and/or powerline communications capability function done by the hub (not the cloud)

not sure about unix, but the rules of multitasking operating systems prohibit direct input/output by a program - programs issue an i/o request to the operating system and the operating system returns the results to the program - an i/o  request can be for data that resides on a 'hard drive' or a print request or whatever - the operating system controls a variety of devices on behalf of itself and programs - a program can construct a record to be written to the hard drive, but it passes that data and request to the operating system

the operating system has hardware and 'drivers' for the devices it communicates with

the 'links table' (i am told) are in the plm - nothing magical there - the isy has its own list of links - the insteon commands and packet structures are published - we need a device and driver for the operating system to issue the packets to the devices - either on the power line (doable - others have done it) or rf (broadcast/receive also doable)

but doing all that does not get over the legal hurdle

 

Insteon is a mesh network.  All the nodes connect to as many other nodes as possible in a non-hierarchical format. 

The hub is a plm with a dedicated controller built-in.  Somewhere on here someone published that they were able "hot wire" a hub to an RS-232 adapter and use it as a standard plm. 

Posted (edited)

Insteon is a little more than the X10 standard. It uses links to see if the sender and receiver know each other. Later ones have the I2CS (I2 extended commands and a check sum). Starts 800 us before the zero crossing.

Now that Insteon is gone from the web. Can't find the clear Insteon Details Whitepaper on the web. I did attached it here

 

insteon_details.pdf

Edited by Brian H
Found paper
Posted
1 hour ago, apostolakisl said:

Insteon is a mesh network.  All the nodes connect to as many other nodes as possible in a non-hierarchical format. 

The hub is a plm with a dedicated controller built-in.  Somewhere on here someone published that they were able "hot wire" a hub to an RS-232 adapter and use it as a standard plm. 

i never got the difference between the plc/plm/serial/usb stuff - never had a need to but here, i use plm - i thought the functions were essentially the same

the underlying physical topology is not a mesh

thinking about it - its more ethernet on the legs - hub from the power panel

its probably not necessary to understand the underlying physical topology - a logical view is enough to work with the protocol commands and addressing 

not sure why people think 'mesh' is a superior topology - the superior topology is one that works and works reliably - insteon devices can and do periodically miss a command - its probably an acceptable tradeoff - adding reliability usually means complexity and larger packets and more code

Posted
1 hour ago, Brian H said:

Insteon is a little more than the X10 standard. It uses links to see if the sender and receiver know each other. Later ones have the I2CS (I2 extended commands and a check sum). Starts 800 us before the zero crossing.

Now that Insteon is gone from the web. Can't find the clear Insteon Details Whitepaper on the web. I did attached it here

 

insteon_details.pdf 1.52 MB · 3 downloads

 

1 hour ago, Techman said:

 

Here's another Insteon paper, comparing various protocols, dated 2012

Insteon compared white paper 2012.pdf 1.29 MB · 1 download

interesting - i never took the time to delve in too deeply - and was not in that development group

i see the 'links' as device associations - no persistent sessions or session awareness - not sure why a device needs a link to another device - or why it would need to be aware of another device - participating in a scene as a controller, it needs to send commands to device addresses - not sure if a link is involved in that - i should take some time and read the manuals posted

 

Posted

So aren't we just talking about a broadcast protocol and individual devices that maintain a table of broadcasts they are supposed to repond to and how? As far as I know there are no sessions and device "awareness" is nothing more than the list of addresses and broadcast commands that a given device has stored in its table to listen for. This has a few advantages:

The tables can be built without a central controller using device-to-device association (button presses)

If there is a controller the various associations and actions still work when that controller is down.

Many devices can respond to a single broadcast so group actions are smooth

Replacing or restoring the controller gateway (PLM) is not the the dramatic, difficult, nightmare that it seems to be with other technologies.

You can repair/restore the tables in devices in a couple of minutes... no waiting to "rediscover" neighbors or "re-optimize" routing info.

When you program everything you are done. No waiting for the mesh to settle or heal or whatever.

If you know the address of a device you can enter it remotely to add it to a system. No crawling into the attic to push the pairing button to initiate the link (though you still have the option to do it that way if you want to)

A broadcast that is repeated by all devices is simpler and naturally resilient to changes in device topography. A single key device can really screw up a routed protocol if it fails or is removed and the recovery time is never as fast or smooth as the product brochure promises it will be.

Posted
17 hours ago, RPerrault said:

i never got the difference between the plc/plm/serial/usb stuff - never had a need to but here, i use plm - i thought the functions were essentially the same

the underlying physical topology is not a mesh

thinking about it - its more ethernet on the legs - hub from the power panel

its probably not necessary to understand the underlying physical topology - a logical view is enough to work with the protocol commands and addressing 

not sure why people think 'mesh' is a superior topology - the superior topology is one that works and works reliably - insteon devices can and do periodically miss a command - its probably an acceptable tradeoff - adding reliability usually means complexity and larger packets and more code

The comm topology is a mesh.  All devices are equal and behave the same as far as comm (except I suppose battery devices).  There are no switches or routers or any pre-defined pathways.  There is no Insteon device that has a role in traffic distribution that is any different than any other device (battery things again).  It is not like ethernet where a switch whose sole job moves traffic according to its destination.  Nor is it like an ethernet hub that divides bandwidth among its outputs.  Insteon has no direction to send traffic, everything goes out in all directions and everyone that hears that message propagates it until it hits the the max hops.  The device that is sold as a "hub" is not a hub in any communication sense.  It plays no different role than any other device with regards to communication.  It is a hub only in the sense of how its user interface presents information to the user.  It looks like a hub to the user, but it isn't.  Calling your circuit panel a hub is not a very accurate description, especially since Insteon works by radio as much as PLC (probably more I would guess). A formal definition of a hub would need to be laid out, but I have never thought of a circuit panel as an Insteon hub and I'm not sure I have ever heard anyone else refer to it as a hub.  When I think of a hub, I think of one client per spoke, but with Insteon, you may have multiple clients on the same circuit and none on any other circuit.  In that case, would you call the gang box where the wires are spliced a "hub" now?  I would say that a better description of these places as "distribution points", the same way an electrician would describe it from the standpoint of electricity.  But again, radio bypasses all of that.

Guest
This topic is now closed to further replies.

×
×
  • Create New...